Most investigators are agreed that acute and chronic arteriovenous fistulae produce an increase in cardiac output (1, 2). Furthermore, a chronic fistula of sufficient size produces an increase of the blood volume (1, 2) and, upon closure of such a chronic fistula, a diuresis of sodium and water (3). On the other hand, it has been shown recently that an acute arteriovenous fistula accommodating 20 per cent or more of the original cardiac output increases the cardiac output but leads to a decrease in the blood flow through the normal systemic capillary bed due to vasoconstriction (1). It appeared of interest, therefore, to determine whether the renal vascular bed participates in this vasoconstriction and, if so, whether an acute arteriovenous fistula affects the renal functions in such a manner as to result in sodium and water retention. METHODSTen adult female dogs were anesthetized by the intravenous administration of 0.47 to 0.57 cc. per kilogram of a 6 per cent solution of sodium pentobarbital. The trachea was cannulated with a Y-shaped glass tube. Segments of the right common carotid and right brachial arteries, right jugular vein, as well as both femoral arteries and veins were dissected free. After the necessary dissection was completed, a multi-eyed catheter was introduced into the urinary bladder. The animal was then given intravenously 5 to 10 mg. of heparin per kilogram after which the vessel cannulations and the other necessary procedures were carried out as described below. The mean arterial blood pressure was recorded with a mercury manometer connected to the central end of the left femoral artery. In 3 of the 10 experiments, the cardiac output was measured according to the Fick prin-

In patients with aortic insufficiency and without disease of the coronary ostia or arteries, the occurrence of chest pain having some or many of the characteristics of cardiac pain due to myocardial ischemia has been ascribed classically to a decrease in the coronary blood flow (1, 2), the aortic insufficiency presumably decreasing the coronary blood flow by lowering the mean aortic diastolic pressure. However, the data available on the effect of aortic insufficiency on the coronary blood flow are contradictory. From acute experiments on the anesthestized dog, Green (3), as well as Green and Gregg (4), concluded that when the aortic insufficiency is marked enough to lower the aortic diastolic pressure, the increase in coronary blood flow occurring during systole does not compensate for the decrease in flow during diastole and the mean coronary flow decreases. On the other hand, Foltz, Wendel, and West (5), from measurements made on anesthetized dogs in which one or two aortic cusps had been torn three or more days previously, concluded that the coronary blood flow was increased over levels usually observed in normal dogs while oxygen consumption was greatly increased over normal levels. Because of the discrepancy in the opinions about the effect of aortic insufficiency on the coronary blood flow, it was thought advisable to reinvestigate this problem. METHODSEight dogs weighing between 17 and 40 kilograms were anesthetized by the intravenous administration of 1 This work was made possible by grants-in-aid from the New York Heart Association, the Sidney A. Legendre Gift and the Charles A. Frueauff Gift.2 Postdoctorate Fellow of the United States Public Health Service.$Fellow of the Dazian Foundation for Medical Research. 4 Research Fellow of the Hudson County (N. J.) Heart Association. 10 ml. per kilogram of weight of a 1 per cent chloralose solution. In the first four dogs the trachea, both common carotid arteries and both external jugular veins were exposed through a midline incision and a Y-shaped glass tube inserted into the trachea. Under artificial respiration, the left hemithorax was opened, the pericardium incised and the heart suspended in a pericardial cradle. The method used to measure and continuously record the output of the left ventricle was a modification of that previously described in this laboratory (6). The proximal ends of the left subclavian artery and brachiocephalic trunk were cannulated. Then the aorta was clamped distally to the subclavian artery so that the blood ejected by the left ventricle was directed into an electromagnetic rotameter (7) and returned to the circulation via both carotid arteries cannulated distally as well as both femoral arteries cannulated proximally and distally. Since the coronary sinus drains blood essentially only from the left coronary artery (8, 9) and since, in a given animal, the coronary sinus blood flow represents a constant portion of the left coronary artery flow (8), measurement of the coronary sinus blood flow was used as a means of studying the left...